
прагматика и медиа дискурс / Teun A van Dijk - News Analysis
.pdf
3. RACISM AND THE PRESS |
185 |
editor Kelvin McKenzie about Blacks in South Africa and the United Kingdom alike. This suggests that the barely veiled racism of the British (or German, such as Bild) tabloids is systematically related to the attitudes of their editors, i.e., leading members of a highly influential elite, controlling vast resources of symbolic media power, and not simply a reflection of what the public wants to read (see also Gormley, 1975). With some well-known transformations (of mitigation, indirectness, or pseudoargumentation), the same holds for the quality press, both in the United Kingdom and in the Netherlands.
The thematic picture for Germany is pardy parallel to that in Britain and the Netherlands (Ruhrmann & Kollmer, 1984). Here, too, the major topics belong to the dominant subjects of crime, residence status (immigration), politics, and identity or integration (yace relations in Britain). The emphasis in these topics is on the use or abuse by foreigners (mostly Turkish) of German resources (welfare, employment, housing, education). Much as in the Netherlands and Britain, another thematic dimension of diese topics is that of what is called `Überfremdung' in German (literally, overalienation), which has also dominated the political and media discussion in France (see e.g., the papers collected in Mots, 8, 1984), Switzerland (Ebel & Fiala, 1983), Sweden (Hedman, 1985), and other Western European countries.
Roles of Minorities
Topics account for global events and situations in which both minority and autochthonous groups are involved. Further analysis of the more specific roles such groups and their members play in the situations accounted for by the media is necessary, however. We have few explicit data for such an analysis, but the topics or themes that were derived from extant content analyses may be further inspected to derive, tentatively, the major roles played by minorities in the news.
For the dominant crime topics, this role is straightforward: Minorities are characterized primarily as agents or suspects of a variety of crimes and very seldom as victims of white crime. Especially in the quality press, this association with crime may be more indirect. In that case, they are involved in drugs, whether as active dealers or as more passive, but still self-inflicted victims of drug abuse. For readers' superficial interpretation, if monitored by ethnic prejudices, there is no difference in this role assignment: In both cases crime or drugs are attributed to minorities, and hence seen as actively participated in as negatively-valued agents (Graber, 1980).
The same is true of the other topics summarized by the overall implication "They cause problems and conflicts." Whether in immigration, housing, unemployment, or education, minorities (or simply minority presence

186 |
PROPERTIES OF NEWS ABOUT ETHNIC MINORITY GROUPS |
or immigration) are implicitly or explicitly represented as the causes (agents) of increasing problems, not as their victims (patients). Similarly, in our thematic analysis, we have shown that minorities often appear as people who are dissatisfied, as protesters, demonstrators, or even rioters, which again identifies the agentive role with negatively valued concepts. They are portrayed as ungrateful dissidents, not as people who rightfully resist against unjust treatment. For the German regional press, Ruhrmann & Kollmer (1984) also conclude that in most cases (more than 50%) foreigners are valued negatively in the press, mostly by the newspapers themselves, by public opinion, or by the government.
Minorities also tend to appear preferably as groups and seldom as prominent individuals as would be the case in typically personalized white news. Their leaders do not receive VIP treatment and are, therefore, barely known by the public at large. If they do appear as individuals, they are mostly criminals or victims. And as we shall show shortly, they are also quoted less often than autochthonous actors.
The role of the autochthonous actors is ambiguous. In much coverage, the state institutions (the government, parliament, city councils and other national and local authorities, the police and the courts, education and research) are portrayed most frequently in ethnic news reports: 70% of all autochthonous actors according to our Dutch data of 1981, whereas the state appears as actor in more than half and the judiciary and the police in
more than a third, of all reports examined from the 1985-1986 data. These actors appear in active roles: They are expected to control immigration; prevent and prosecute crime; or to provide assistance in employment, housing, or education. They are expected to analyze and solee problems and to develop the best policies to control the -minority problem". Ruhrmann Kollmer (1984) showed that for Germany the government and public opinion are most often (about 80%) mentioned as the actors from which action is expected or demanded or to which recommendations are addressed.
In the Dutch data of 1981, the police and the courts are mentioned even more frequently than the national or local government. The role of the authorities is characterized mostly in neutral terms: The actions described are the normal task or duties of the authorities. Sometimes, especially in the conservative British press, the authorities are criticized for not being strict enough or for giving priority treatment to Black immigrants. We found that the government and the police are especially represented more often in negative terms than other institutions. But even when criticized of "not doing anything against it", the active role of the authorities is always presupposed. The characterization of government is more positive when it is portrayed as the Great Helper—as the dispenser of goods, money, and services for the minority groups. The frequent reference to protest and dissatisfac-

3. RACISM AND THE PRESS |
187 |
tion by the ethnic groups puts this role in further perspective, and rhetorically enhances it by implied contrast.
Other authochthonous groups are represented in a variety of roles. Individual Dutch people and employers usually appear as agents in discriminadon news. Members of parliament occur as either supportive of restrictive government policies or as opposed to them, as may be expected, and the same is true for political parties. When supportive of immigration and immigrants, or when acting as antiracists, they are portrayed as active but negative agents (see also Seidel, 1987b). But since they are ingroup members, their actions of protest, dissent, and demonstration are implicitly viewed as racial treason and dealt with in even more negative terms than those of the outgroup members.
On the other end of the scale, members of racist parties, people who discriminate, and people who oppose immigration or the special provisions for minorities are usually represented in an active role. This time, however, while superficially condemning right-wing action, the conservative media represents it within a context of reaction, that is, as partly understandable or even excusable: Racists are provoked to act against minorities. The disco or bar bouncer who does not admit more Blacks or the employer who fires a Turkish worker because of problems (s)he has had with him or her is portrayed as acting out of ignorance because of an especially difficult situation, or because he or she is an incidental bigot. Such forms of everyday racism (Essed, 1984) are seldom placed within a wider perspective of structural racism in society. At best, active discrimination is represented as outside the liberal consensus but never as a serious crime. (Like speeding or drunk driving; one shouldn't do it, but everybody does it).
Thus, for much of the dominant press, right-wing racists are seen as active deviants, who have placed themselves outside of the broad ethnic consensus of apparent liberalism and conditional tolerance. They are not so much characterized as active, responsible agents of racist policies and actions as they are illegitimate competitors in the political arena. They steal the votes of the frustrated white poor and say embarrassing things, even when they are sometimes right when they claim priority treatment for "our own people."
At the local level of syntactic organization, role distribution has been systematically investigated by Fowler and his associates (Fowler, et al., 1979; Fowler, 1987). They show that the action structure of sentences, as expressed by word order, grammatical relations (subject, object, etc.), actives, and passives in news reports in the British press about disturbances also signa' that minority groups or their members tend to show up as agents, especially in negative contexts, whereas the police will be put in a less direct passive phrase or remain implicit altogether. Sykes (1985, 1987) uses the

188 |
PROPERTIES OF NEWS ABOUT ETHNIC MINORITY GROUPS |
same analysis to show this biased perspective in official agency reports, the syntax of which subtly signals that minorities have no positive and active control of their own situation.
Finally, in the thematic analysis of D2 we examined which actors tend to appear together in the same news item. This association is important in the formation of situation models by readers. They will want to know which typical participants are involved in each situation.. Minorities in general mostly appear together with the national or local authorities or with the judiciary and the police. Surinamese are portrayed most often together with the judiciary and the police, which confirms the criminalization of this group of Black citizens. This is even more pronounced for Turks and especially Moroccans who appear with the police or the judiciary in more than 20% of the items (as immigrant workers or foreigners, they also often appear together). Tamils and Moluccans are mostly confronted with the government, the city councils, or other authorities. The cities have to deal mostly with minorities in general or with the government or ministries, usually when having conflicts about finding minority programs or provisions. The judiciary, apart from acting with the police, appear to deal mostly with Moroccans, expecially in decisions about, or legal actions against, expulsion of illegals. The police, finally, is mostly associated with minorities in general, and with Turkish people, either because of crime or again because of ilegal
presence in the country.
These associations give further substance to the analysis of role relationships in the thematic structures that organize the global contents of news reports about minority groups. They confirm our repeated finding that press stories about minorities are not so much descriptions of a large variety of possible news events but rather an expression of stereotypical situations to which minority presence is restricted: immigration and expulsion, discrimination, demands, dissent, social and cultural problems (language, education, housing, and employment), and finally crime and deviance. Other topics and other roles are virtually absent, such as their own political and social organization, health, culture and the arts, and especially experiences of racism. Even more than for news in general, reporting about minorities appears to be both stereotypical and negative, so that some topics (politics, economy, culture) dealt with when autochthonous groups are actors are seldom prominent in ethnic news.
Headlines
A special study of all headlines in the 1985 data (van Dijk, 1987d) confirms most of the previous conclusions. Headlines are particularly important because, both in production and in the reception of news reports, they subjectively define the most prominent or most relevant information of the news

3. RACISM AND 'THE PRESS |
189 |
item. They express—sometimes in a biased way—the top of the underlying semantic macrostructure, and at the same time define the situation reponed in the press. Since headlines and leads are often the only information read or memorized, they play an important role in further information processing and possible effects of news about ethnic minority groups.
As may be expected from our subject and topic analyses, immigration is the subject most often (17.6%) expressed in the headlines, followed by crime (12.8%) and discrimination (11.1%). In this respect, the headlines correspond more or less with the themes they summarize, except for crime which now occupies second position, suggesting the well-known operation of relevance upranking. Several secondary topics follow (each about 2% or 3%), such as "action groups helps minorities", -minorities participate in elections", or "minorities make demands or protest." Interestingly, only 0.6% of the headlines feature the subject of white violence against minorities.
Over half of all headlines explicitly mention as actors ethnic minorities, refugees, or specific ethnic groups. Refugees (Iranians, Tamils, etc.) are especially made prominent in the headlines (see also Chapter 4). Together with the extant groups—Surinamese, Turks, Moroccans, and Moluccansthey now count among the most frequent headline actors in the Dutch press.
White institutions and groups, such as the government, the cities, the judiciary, or the police appear relatively infrequently in the headlines (together in only about 10% of the headlines). That is, although they appear as actors in most reports, their role apparently remains downgraded or implicit in the headlines. In other words, the ethnic event is defined as involving ethnic groups, even when the Dutch authorities play a prominent role in such events.
These role relationships become particularly clear when we examine the syntactic position and semantic roles of actors in the headlines. This analysis shows that in more than 25% of the headlines, minority groups appear in first sentence topical position. This position is usually reserved for active agents in Dutch and English syntax. Not so in these headlines, where the minorities in topical position have active agent roles in only 7.7% of the headlines. It follows that they mostly are experiencer, or object of actions by others. Since these others must be the Dutch authorities and since these do not frequently occur in the headlines, we conclude that most headlines involve such authorities only in an indirect or implicit way, as e.g., in such actions as expel, refuse, arrest, etc. In the few headlines in which minorities appear as active agents, they are mostly agents of negative or negatively associated actions, such as demanding, protesting, resisting, going to court, complaining, or engaging in a hunger strike. The association with crime in the frequent crime headlines is more indirect and becomes explicit es-

190 PROPERTIES OF NEWS ABOUT ETHNIC MINORITY GROUPS
pecially when minority involvement can be defensibly associated with cultural habits, such as the prominent headlines in the national press about a Moroccan father who circumcised his child with a pair of scissors. There are no straightforward positive agent roles for minorities in the headlines.
These results are confirmed by an analysis of the headline roles of the white authorities: If they appear explicitly as first topics in the headlines at it is most often as neutral (typical actions, such as arresting, dislodging, etc.) or positive (throwing a party for minorities in the police station, learning Turkish, and generally wanting to ameliorate ethnic relations) agent
roles.
This headline analysis suggests that (1) special topics such as immigration and crime tend to become further emphasized; (2) ethnic situations primarily involve ethnic groups, although white groups or authorities in fact are more prominent in such news; (3) minorities often have first-position, topic role in the headlines, but are seldom agents; if they are agents it is usually in more or less negative contexts; and (4) white groups or institutions occur much less in the headlines (at least not explicitly), but if they occur in topic position, then usually are neutral or positive agents.,
Four years earlier, the headlines in the Dutch press appeared to be similar. Except for De Telegraaf, headlines no longer explicitly associate minority groups with crime (even if crime is a major subject in headlines), but this does not mean that they are always neutral. NRC-Handelsblad defines the situation in a Moluccan settlement with a big headline such as
MOUNTING TENSION IN SETTLEMENT, and in another item in the same newspaper rather than a headline that summarizes the main event—"Government Stops Subsidy. - On the other hand, negative actions of the authorities are seldom headlined. In the case of a racist pamphlet distributed after a trial of a foreign murderer, De Volkskrant headlines PROTESTS AFTER JUDGE'S VERDICT IN STABBING CASE. In other words, the protest nature of the pamphlet is defined as the most important dimension, not the fact that the pamphlet was racist or that the police helped distribute it. Similarly, after the squatter demonstration analyzed in Chapter 5 this same newspaper uses the headlines RIOT AFTER EVACUATION and HOOLIGANS STILL UNDER ARREST, although most of the article is about the treatment of the Surinamese families in the police station. No headlines, however, appear such
let alone

3. RACISM AND THE PRESS |
191 |
less negative than Telegraaf style headlines of the type Tumc, 16, GETS SLX
YEARS FOR MURDER.
Local Meaning, Style, and Perspective
The informal thematic and role analysis given aboye already suggests a number of impIications for the study of style and perspective. Given a negative topic, we may expect that local meanings and their expression in lexical items and syntax will also match the overall topical orientation and perspective. If minorities are represented as responsible agents in a host of problems, this should also surface in their person and action descriptions and in the perspective conveyed by sentence syntax. This is indeed the case, but we have as yet little systematic data to sketch the full picture of this local dimension of ethnic coverage. This generally holds for qualitative analyses, which were neglected or impressionistic in most quantitative studies.
Who is Speaking?
One important clue about the perspective of the media on ethnic affairs comes from an analysis of speaker roles. When and how often are minorities quoted or referred to as speakers, as people who voice facts and opinions? Downing (1980) found that minority group members, like Black African leaders, are quoted less often than white spokespersons, even in accounts of events that directly concem them. Similarly, in our 1981 press data we found that of all speakers in the minority news, 70% are white autochthonous Dutch (mostly institutional spokespersons, such as politicians), despite the fact that in most of the news reports (61%) minorities were major actors. The 1985-1986 data show a similar picture. Majority group actors are quoted in nearly half of all occurrences but minority actors are quoted only 25% of the time.
This bias in the distribution of speaking roles has severa! structural and cognitive causes. First, as dominated groups, minorities in Western Europe are less organized and, therefore, have less organized access to the media through press conferences, press releases, or designated spokespersons. Second, for the same organizational reasons, journalists are less inclined to actively search for or listen to minority sources. Third, such sources are perceived as less objective and hence less credible, which also is manifested by the predominant use of words that express doubt and distance and the more explicit use of quotation marks when they are allowed to speak. As interested party they are seldom or cautiously quoted when they accuse Dutch people or institutions (e.g., the police) of discrimination or racism. This is much less the case when the opinions of the authorities are men-

192 |
PROPERTIES OF NEWS ABOUT ETHNIC MINORITY GROUPS |
tioned, for instance, when they deny discriminatory acts. Fourth, practically all journalists in Western Europe are white and have no personal or professional relationships with minority groups. Due to their subtle or more blatant prejudices or to the usual problems of intercultural communication, white journalists will feel much less comfortable when interviewing Black spokespersons and will, therefbre, also tend to avoid them as sources. Fifth, the converse reinforces this tendency: Since most credible sources—the authorities, politicians, educators, professionals, or scholars—are white and since most institutions are white, they are more likely to be used as sources.
U.S. research shows the same tendency. In general, but especially when ethnic or racial conflicts arise, the version of events provided by the authorities, most notably the police, prevails. Gutiérrez (1978) showed that not only Chicanos tend to be represented stereotypically by the Californian press but also that their point of view of the conflicts in which immigrant workers are involved is subordinated to that of the Anglo officials. This has also been a permanent observation about the press accounts of Black riots in the 1960s and before (Fischer & Lowenstein, 1967; Knopf, 1975). The same is true today for the coverage of the riots in the British cities during the 1980s (Downing, 1985; Murray, 1986; Sivanandan, 1986; Tumber, 1982).
Evaluation, Implication and Presupposition
More than other news, the coverage of ethnic minorities has an evaluative dimension: Opinions are more often expressed or implied than for instance in political news (Ruhrmann & Kollmer, 1984). In our own analysis of Dutch news, we first found that the national newspapers did not resort to explicit racial slurs, except in quotes describing the words of racist people. Negative evaluations are usually more subtle and indirect. Most consistent is the negative dimension implied by the set of words designating problems, difficulties, and conflicts associated with ethnic affairs. This negative aspect may he further emphasized by the use of words such as "threat" even in contexts where it means "to intend" or "to announce": Minorities are described as "threatening" with legal action or a demonstration, whereas the authorities never "threaten" to stop fmancial help, but "decide" or "announce" to do so. This usage is coherent with the more general lexical framework of violence terms associated with minority action and expresses prominent underlying prejudices about minority groups, especially Black men. Ruhrmann & Kollmer (1984) show that negative evaluations in the German regional press are usually framed in terms of criticism, deviation of the norm, and stereoiypes.
A key property of discourse semantics is implicitness. Only part of the information must be expressed in the text itself. Writers can leave most inferences to the reader. In ethnic news, this means that negative implica-

3. RACISM AND THE PRÉSS |
193 |
tions or associations need not be expressed. In fact, in many cases such implications are not directly obvious to an uncritical reader. They only appear as further inferences are drawn or when several examples are compared to yield cumulative associations. The following examples are from our 1981 data:
•(A Turkish woman looses her residence pelma) "which she lost when
she ran away from her husband because of assault" (Volkskrant, Oct. 1).
•(Because of the city action plan against drugs) "A [Surinamese] welfare
agency fears occupation . . . . The agencies have been threatened ..... (Volkskrant, Oct. 7).
•anarchistic situation in the camps— (Volkskrant, Oct. 7).
•"Half of the foreign children have been in contact with the police . . . their fathers sometimes give them quite a beating ... or the children are threatened with death" (Volkskrant, Oct. 8).
•( Moluccan drug addicts in Groningen, alter a subsidy was stopped for their center) ". . . windows have been smashed ... there has been word of hijackings . . . and it is feared that the stones will be followed by more serious eruptions" (NRC-Handelsblad, Oct. 2).
•(An editorial about the drugs scene:) "harassment by the heroin traffic . traumatic episode of the drugs cafés, ... a center for Surinamese addicts and its ilegal follow-up, . . . tribal war among the social workers . . ." (NRC-Handelsblad, Oct. 9).
•(About P. the leader of the Moluccan settlement group:) Mayor is reported saying "they cannot be trusted", and "P. speaks a mixture of formal Dutch and legal jargon." The mayor: "P. is nearly criminal, and misleads his own people" (NRC-Handelsblad, Oct. 9).
•(Interview with Moluccan leader P.) "piercing looks, a long story . . . from which it should be inferred ... he openly speaks about violent resistance ... He threatens to . . . People who carne to collect
the rent have fear . . . the beginning of terror . . |
(Het Parool, Oct. |
6). |
|
•(About the departure of gypsies:) "others try to enter Germany illegally, via backroads" (Het Parool, Oct. 7).
•"With eleven knife blows, Z. . . killed the . . . (De Telegraaf, Oct. 1)
•(About a demonstration of Surinamese before the town hall:) "the demonstrators were mostly divided . . . there was much shouting
. . . that nobody understood" (De Telegraaf, Oct 6)
•A shooting in front of a café "where a lot of foreigners used to come"
(De Telegraaf; Oct. 13).

194PROPERTIES OF NEWS ABOUT ETHNIC MINORITY GROUPS
•(A neglected child, mother arrested:) "neglection ... drugs ... heroin traffic ... prison . ." (De Telegraaf, Oct. 31).
This is merely a small sample of illustrations. Unfortunately the English versions lose much of the subtleties of the original Dutch. Yet, the major implications are obvious and confirm the general topics examined before: Ethnic groups are violent, threatening, cannot be trusted, are funny or strange, do not speak our language properly, neglect their children, are involved in the drugs scene, and are generally threatening majority-group members or the organizations set up to deal with them. If a Turkish woman is described to lose her residence permit because of divorce from her husband, all newspapers briefly indicate the important, but in this context, irrelevant reason that her man was violent. This would not happen when the reasons for a white Dutch woman's divorce were the issue. In fact, violence against women is one of those topics that gets marginal treatment in the male-dominated Dutch press.
This touches upon an interesting feature of text and talk about ethnic groups. In our fieldwork data (van Dijk, 1984a, 1987a), we found frequent mention that Islamic men oppress their women. This generalizing evaluation suggests a feeling of supe riority of Dutch men with respect to backward foreign men, as well as similar feelings of white women towards their less emancipated sisters (Essed, 1983). The faz-from-ideal position of Dutch women in the Netherlands, despite a decade of feminist struggle, seems conveniently overlooked in such derogatory discourse. Thus, the true or alleged sexism of foreign men has become a preferred, while safe stereotype (indeed, what liberal could object to such a charge of sexism?) also in the press. Yet, its frequent expression suggests that it is part of a more complex negative attitude towards foreigners, rather than of a positive attitude towards women. Similar conclusions may be drawn from the examples about the family life styles of Turks or Moroccans, the role of fathers and sons, and the problems this creates for the Dutch authorities.
Even more subtle are those implications presupposed by the newspaper or the person(s) interviewed. Often they pertain to assumed properties of ethnic groups:
•(An interview with an official) "These people are sometimes very intelligent" (Volkskrant, Oct. 7).
•"They said they would keep their promise" (NRC-Handelsblad, Oct. 9).
These are two examples that presuppose that an ethnic group is usually not so intelligent and that usually they do not keep their promises.